Short Report Spawning habitat in the spined loach, Cobitis taenia (Cypriniformes: Cobitidae) Jörg Bohlen Czech Academy of Sciences, Institute of Animal Physiology and Genetics, 277 21 Libe˘chov, Czech Republic (e-mail: [email protected]) Received: December 21, 2001 / Revised: October 2, 2002 / Accepted: November 20, 2002 Abstract Spawning habitat of the endangered spined loach Cobitis taenia was studied in northwest- Ichthyological ern Germany. The distribution of eggs of the loach in the field was best correlated with the density of Research vegetation but had little correlation with current velocity, water depth, or bottom substrate. In the aquarium, the loach chose dense vegetation for an oviposition site, as inferred by the direct observa- ©The Ichthyological Society of Japan 2003 tions of spawning and the location of spawned eggs. It is concluded that such habitat specialization is an important element in the autecology of the endangered fish and a critical stage in the habitat Ichthyol Res (2003) 50: 98–101 requirements of its populations. Key words Cobitis taenia · Spawning substrate · Habitat preference · Reproductive biology pawning in many fish species includes behavioral ele- level of the creek channel. The bottom mainly consisted Sments that ensure suitable conditions for eggs and of sand, often with addition of silt but rarely of gravel or young after hatching. A common behavioral element is the peat. Vegetation covered about 30% of the bottom and selection of a suitable location through preference of cer- was dominated by spike rush, Eleocharis acicularis, cover- tain spawning structures (Balon, 1975). The suitability of the ing contiguous areas of several square meters. Callitriche selected location for egg and larval development influences palustris and fusiform algae (Chladophora) together mortality of the offspring, especially in species without formed more or less dense bunches, usually not larger than parental care. 300cm2. The floating underwater leaves of Sparganium The spined loach, Cobitis taenia Linne, is listed as an emersum covered as much as 1000cm2. Sometimes terres- endangered species in many European countries (Kotusz, trial grass had fallen into the water, and the roots formed 1996), but the reasons for its decline are uncertain (Lelek, very dense bunches. The vegetation types were distributed 1987). This fish is highly specialized for life on and in sandy in a patchy pattern mainly along the shallow edges. The bottoms. During the daytime, the fish burrows in the sand, described conditions can be regarded as typical for the sum- and during the night it feeds on the sand by a specialized mer situation. Haaren Creek is believed to house a pure filter-feeding mechanism (Robotham, 1982). Further population of Cobitis taenia (Bohlen et al., 2002). habitat requirements other than sandy bottom have not Lake Müggelsee is located in the drainage of River Elbe been reported in this fish until now (Lelek, 1987; Blohm at Berlin, eastern Germany. The eutrophic lake has a surface et al., 1994). area of 7.3km2 and an average depth of 4.9m (Driescher The aim of the present study was to estimate habitat et al., 1993). Shores are shallow and partly vegetated selection for oviposition in the spined loach by identifying by Phragmites sp. Submerged vegetation is restricted to its preferred spawning substrate in the field and in aquarium filamentous algae (Chladophora), but a summer bloom of experiments. Cyanobacteria occurs regularly. The bottom consists of a varying mixture of mud, sand, and gravel with a layer of fine to coarse detritus along the shore. A complex population Materials and Methods including Cobitis taenia plus at least three hybrid forms between C. taenia and C. elongatoides lives in Lake Study site.—Haaren Creek (Weser River catchment area, Müggelsee (Bohlen et al., 2002). northwestern Germany) at the sampling locality 3km west Field study.—Eggs were collected on 5 July 1997 in of Oldenburg is about 3m wide, and its bottom is about Haaren Creek by placing a plastic tube onto the creek bed, 1.2m below the level of the surrounding cattle grassland. At thus isolating an area of 529cm2. In a 150-m section of the the sampling stretch, maximum depth was 0.25m. Maximum creek, 85 samples were taken at randomly chosen points. water level during 1995 to 1997 always occurred in February The water in the tube and the top layer of the bottom were and reached a maximum monthly mean of 90cm. From thoroughly sieved with a net with 1-mm mesh. For each of spring to autumn, the water level does not reach half the the sampling points, the total number of eggs was noted and Spawning substrate in Cobitis taenia 99 the habitat parameters were estimated. The presence of eggs in a sampling point was tested for the effects of habitat parameters by a multiway analysis of variance (ANOVA). In Lake Müggelsee, eggs were collected nonquantitatively by pulling a fine-meshed net along the bottom or by wash- ing vegetation and detritus in a bucket. Laboratory study.—Spined loaches from Haaren Creek were bred in aquaria during 1998 as described in Bohlen (1999). Each aquarium measured 39 ϫ 39 ϫ 25 (h) cm and was equipped with a layer of fine sand, an air-driven sponge filter, and a plastic box (20 ϫ 20 ϫ 5cm) covered with metal gauze (mesh size, 5mm). Cryptocoryne affinis and Sagittaria sp. were set on the bottom (about 20 plants per tank), Elo- dea canadensis floated at the surface (per tank, 2–5 trunks of 10–15cm length), and a bunch of moss (Vesicularia dubyana) of approximately 150–200ml was located on the top of the plastic box. Eggs spawned into the moss fell into the plastic box and were removed and counted. To find scattered eggs on the bottom, a fine-meshed net was used. One to five pairs of fish per aquarium were adapted to the experimental conditions for at least 5 months before the first spawning. Spawning behavior was observed on two occasions. At one occasion, the claspings of the female by the males with release of eggs were counted for the sub- strate selection. Results Field study. Eggs were found in 17 of the 85 sampling points (20%). The number of eggs at the 17 points ranged from 1 to 44 with a mean of 7.8 and a total of 133 eggs. Percentage of sampling points containing eggs and the percentage of eggs per category are given in Fig. 1. No significant correlation between the number of eggs per sam- pling point and any habitat parameter was found. The pres- ence or absence of eggs, however, was explained by the habitat parameter type of vegetation (F ϭ 2.905, P ϭ 0.02). No significant influence as a source of variation was ob- served in the parameters current speed (F ϭ 1.095, P ϭ 0.36), depth (F ϭ 0.892, P ϭ 0.47), and type of bottom (F ϭ 0.659, P ϭ 0.66). The type of vegetation preferred by the loach was the dense bunches and the grass. Sampling of Cobitis eggs during the summer of 1996 and 1997 in Lake Müggelsee revealed similar results. On ten sampling dates, 80 eggs were found within the dense masses of fusiforme algae (Chladophora), which occurred in water Fig. 1. Frequency (%) of habitat category (white columns), of total egg depths of 25–45cm. No eggs were found outside this struc- number (gray columns), and of sampling points containing eggs (black ture, and no eggs were found after the algae disappeared in columns) within 85 samples from Haaren Creek, Germany, in relation July. No eggs were found in the belt of rough detritus, which to type of vegetation (a), water depth (cm) (b), type of bottom (c), and Ϫ1 indicates a preference of vegetation in water of medium current velocity (in m s ) (d) depth rather than detritus in shallow water by the spawning fish. Laboratory study. From 74 spawning acts, 33 482 and inside the plastic box indicates a strong selection for the 1057 eggs were counted inside and outside the box, respec- spawning substrate. tively (Table 1). Altogether, 96.9% (range among spawning During the observed spawning, 36 claspings with release acts, 62.4%–100.0%) of the eggs were found within the plas- of eggs were counted. Of these, 35 (97.2%) occurred within tic box. Because the surface area of the box was only 20% of the bunch of moss and only 1 occurred between strains of the total surface area available, the high percentage of eggs Elodea. The spawning fish penetrated into the vegetation, 100 J. Bohlen Table 1. Data on the spawning of spined loach in aquaria Number of Number of Total number Number of eggs Percentage of eggs placed females per replicates of spawning per night into the moss mean aquarium acts mean (range) (range) 1646 395 (97–875) 96.3 (62.4–100) 2211 436 (220–1091) 97.8 (91.4–100) 5117 681 (38–1557) 95.1 (68.4–100) Total 9 74 467 (38–1557) 96.3 (62.4–100) Number of females per aquarium, number of replicates, and the number of spawning acts are given The total number of counted eggs per night and the percentage of eggs placed into the moss show the selectivity of the spawning fish for the moss as preferred substrate for oviposition stopped there, and the male embraced the female during can be characterized as a fine, soft, and dense bunch. This egg release. During the second observation, no counts of the preference represents another habitat requirement in addi- claspings were done, but the behavior of the fish was the tion to the requirement of sandy bottom as shelter and same as during the first observation.